I can chime in here in that it would really simplify workflows.  We do RME on a document level, not setting up record series or folders.  It makes more sense to the users, and managing transparent records management is a huge overhead and many people simply don't understand it.

All of our workflows are filled with giant conditional branches that have to check the retention and then assign the right one.  Essentially the same number of branches as all possibilities for retention.

When document are captured we typically look up to a database that houses the document type and retention.  We then check that action on the conditional branches.  It would be so much easier to simply be able to put a token in there and have one branch.  As far as error handling, maybe if the schedule didn't exists just have the workflow error.

Not speaking for the original poster but it looks they have an upload form where they upload the document and select the retention.  Typically we have users pick the document type, import it, and the workflow that files it auto determines the retention based on type but this still doesnt get around having all conditional branches.

Another item we have is usually a form to manage the document types and retention.  So a user can go and add a document type that will then show up in the dynamic fields.  Or edit a document type and change the retention.  This is a form that on submit goes and updates the database, but also goes and finds the existing documents and updates their retention/info to where again you have the giant conditional branches.